Narrow Results

ATRIAS 1.0 is a spring-legged, monopod robot designed and built as a prototype towards a human-scale 3D biped. The monopod has to meet certain requirements concerning locomotion dynamics and energy efficiency to meet the goal of a biped that can autonomously walk and run efficiently and robustly outdoors, untethered over realistic (non-ideal) terrain. The design of ATRIAS 1.0 includes adequate control authority for robust locomotion as well as incorporating the idea of passive dynamics for high energy economy. Towards this effort, the passive dynamics of ATRIAS 1.0 are designed to match the key features of the Spring Loaded Inverted Pendulum model: a massless leg, mass centered at the hip joint, and a series spring between the ground and the mass at the hip joint. In this paper the authors discuss the key features of this unique robot design.

The present work aims to improve students’ interest in music teaching and promote modern teaching. A distributed application system of artificial intelligence gesture interactive robot is designed through deep learning technology and applied to music perception education. First, the user’s gesture instruction data is collected through the double channel convolution neural network (DCCNN). It uses the double-size convolution kernel to extract feature information in the image and collect the video frame’s gesture instruction. Secondly, a two-stream convolutional neural network (two-stream CNN) recognizes the collected gesture instruction data. The spatial and temporal information is extracted from RGB color mode (RGB) images and optical flow images and input into the two-stream CNN to fuse the prediction results of each network as the final detection result. Then, the distributed system used by the interactive robot is introduced. This structure can improve the stability of the interactive systems and reduce the requirements for local hardware performance. Finally, experiments are conducted to test the gesture command acquisition and recognition network, and the performance of the gesture interactive robot in practice. The results indicate that combining convolution kernels of $5\times 5$ and $7\times 7$ can increase the recognition accuracy of DCCNN to 98% and effectively collect gesture instruction data. The gesture recognition accuracy of two-stream CNN after training reaches 90%, higher than the mainstream dynamic gesture recognition algorithm trained with the same data set. Finally, the recognition test of gesture instructions is carried out on the gesture interactive robot reported here. The results show that the recognition accuracy of the gesture interactive robots is more than 90%, meeting the routine interaction needs. Therefore, the interactive gesture robot has good reliability and stability and is applicable to music perception teaching. The research reported here has guiding significance for establishing music teaching with multiple perception modes.

A strategy for working with incomplete information is called competitive if it solves each problem instance at a cost not exceeding the cost of an optimal solution (with full information available), times a constant. This paper strives to demonstrate why competitive strategies are useful for the design of autonomous robots. They guarantee a good worst-case behaviour, they are easy to implement, and they allow to deal with some problems whose optimal solution would be NP-hard. We survey competitive strategies for the following problems. How to find a door in a long wall, how to find a goal in an unknown environment, how to find a point from which an unknown environment is fully visible, and how to determine a robot's location on a known map from local visibility.

A Magnetic Switchable Device (MSD) is a ferromagnetic circuit using permanent magnets where the flux can circulate between different paths when its configuration is changed. This routes or cancels the flux trough specific surfaces, and thus turns on or off adhesion forces. We present classic and innovative magnetic configuration to realize powerful MSD. We designed and prototyped some miniature systems and give their characteristics. Finally various robotics applications for gripper, anchor and climbing robot are unveiled where the MSD solution has proved to be advantageous.

In this paper, a new robot path planning algorithm based on Quantum-inspired Evolutionary Algorithm (QEA) is proposed. QEA is an advanced evolutionary computing scheme with the quantum computing features such as qubits and superposition. It is suitable for solving large scale optimization problems. The proposed QEA algorithm works in the discretized environment, and approximates the optimal robot planing path in a highly computationally efficient fashion. The simulation results indicate that the proposed QEA algorithm is suitable for both complex static and dynamic environment and considerably outperforms the conventional genetic algorithm (GA) for solving the robot path planning problem. Our algorithm runs in only about 2s, which demonstrates that it can well tackle the optimization problem in robot path planning.

As increasingly more research efforts are geared towards creating robots that can teach and interact with children in educational contexts, it has been speculated that endowing robots with artificial empathy may facilitate learning. In this paper, we provide a background to the concept of empathy, and how it factors into learning. We then present our approach to equipping a robotic tutor with several empathic qualities, describing the technical architecture and its components, a map-reading learning scenario developed for an interactive multitouch table, as well as the pedagogical and empathic strategies devised for the robot. We also describe the results of a pilot study comparing the robotic tutor with these empathic qualities against a version of the tutor without them. The pilot study was performed with 26 school children aged 10–11 at their school. Results revealed that children in the test condition indeed rated the robot as more empathic than children in the control condition. Moreover, we explored several related measures, such as relational status and learning effect, yet no other significant differences were found. We further discuss these results and provide insights into future directions.

This paper introduces ATRIAS 2.0, a new platform for the study of bipedal locomotion in robots. One of the purposes of the robot is to further explore the role of compliance in achieving energetically efficient and agile locomotion. A second purpose is to inspire the development and experimental validation of analytical feedback control algorithms for dynamic 3D locomotion. A third purpose is to address the challenge of tightly integrating hardware and software to achieve extreme robustness to unknown ground-height variations while walking or running.

The dynamic modeling and coupling effect of a space robot are complex when the flexible manipulator and solar panels are considered. This paper investigates the dynamic coupling effect and control of a flexible space robot with flexible manipulators and flexible panels. The equations of motion are derived for the robot model both of the rigid-flexible type and flexible-flexible type. The flexible space robot dynamic model is verified by comparison with the results generated by the ADAMS software, for which good agreement has been obtained. The dynamic coupling matrix of the flexible space robot is derived based on the dynamic model. The effects of the central rigid body mass and the joints angle on the dynamic coupling are analyzed. A control method is proposed to manipulate the flexible space robot based on the system dynamic model. The multiple-impulse robust (MIR) input shaper is used to suppress the vibration of flexible structures in the proposed controller. Appropriate design parameter and frequency scaling factor are selected for the MIR input shaper to suppress the flexible vibration. The flexible space robot control is conducted to illustrate the effect of the proposed controller. It is shown that the proposed control method can realize the desired joints manipulation, while suppressing the vibration of the flexible manipulators and flexible panels.

Controlling the locomotion of kinematically complex robots is a challenging task because different control approaches are needed to operate safely and efficiently in changing environments. This paper presents a graph-based behavior description which allows to dynamically replace behaviors on a robotic system. In the proposed approach, every behavior is represented as a directed graph that can be encoded into a data block which can be saved to or loaded from a behavior library. Since this is not a precompiled module like in other systems, the algorithm and parameters of a behavior can still be adapted online by modifying the data that represents the behavior. Thus, machine learning algorithms can optimize an existing behavior to an unknown situation, e.g., a new environment or a motor failure. With a first implementation, it is shown that the proposed behavior graphs are suited for controlling kinematically complex walking machines.

In biological systems, instead of actual encoders at different joints, proprioception signals are acquired through distributed receptive fields. In robotics, a single and accurate sensor output per link (encoder) is commonly used to track the position and the velocity. Interfacing bio-inspired control systems with spiking neural networks emulating the cerebellum with conventional robots is not a straight forward task. Therefore, it is necessary to adapt this one-dimensional measure (encoder output) into a multidimensional space (inputs for a spiking neural network) to connect, for instance, the spiking cerebellar architecture; i.e. a translation from an analog space into a distributed population coding in terms of spikes. This paper analyzes how evolved receptive fields (optimized towards information transmission) can efficiently generate a sensorimotor representation that facilitates its discrimination from other "sensorimotor states". This can be seen as an abstraction of the Cuneate Nucleus (CN) functionality in a robot-arm scenario. We model the CN as a spiking neuron population coding in time according to the response of mechanoreceptors during a multi-joint movement in a robot joint space. An encoding scheme that takes into account the relative spiking time of the signals propagating from peripheral nerve fibers to second-order somatosensory neurons is proposed. Due to the enormous number of possible encodings, we have applied an evolutionary algorithm to evolve the sensory receptive field representation from random to optimized encoding. Following the nature-inspired analogy, evolved configurations have shown to outperform simple hand-tuned configurations and other homogenized configurations based on the solution provided by the optimization engine (evolutionary algorithm). We have used artificial evolutionary engines as the optimization tool to circumvent nonlinearity responses in receptive fields.

Patients who suffer from stroke have motion function disorders. They need rehabilitation training guided by doctors and trainers. Nowadays, robots have been introduced to help the patients regain their motion function in rehabilitation training. In this paper, a novel multi degree of freedom (DOF) exoskeleton robot, with light weight, including (6$+$1) DOFs, named as Rehab-Arm, is proposed and developed for upper limb rehabilitation. The joints of the robot are equipped with micro motors which are capable of actuating each DOF respectively and simultaneously. The medial/lateral rotation of shoulder is realized by a semi-circle guide mechanism for convenience consideration and safety. The robot is used in sitting posture which is attached to a custom made chair. Hence, the robot can be used to assist patients in passive movement with 7 DOFs of the upper limb for rehabilitation. Five adult healthy male subjects participated in the experiment to test the joint movement accuracy of the robot. Finally, subjects can wear Rehab-Arm and move their upper limb, led by micro motors of the robot, to perform task assigned with specific trajectory.

The presence of normal upper limb arm swing movement is considered an important gait movement that affects the selective coordinated locomotor control of the shoulder, elbow, and wrist joint movements. However, in patients with neurological disorders, flexor synergy is characterized by decreased selective neuromuscular control and reciprocal disinhibition of the wrist flexor muscles associated with spasticity or shortness. This research aimed to demonstrate the reliability, validity, and feasibility of the progressive exoskeletal robotic shoulder joint kinematics system. The robotic shoulder joint kinematics system comprises a gait function-retraining robot designed to provide arm swing. The changes in the shoulder joint angle between ImageJ motion analysis software and robotic shoulder joint kinematics system were compared in this research to investigate the reliability and validity of the latter. The linear regression analysis revealed good correlation between the measured angles and the shoulder angle data ($R^2=0.943$). Furthermore, the test–retest reliability test demonstrated excellent reliability ($R^2=0.978$). The robotic shoulder joint kinematics system generated successful arm swing during the locomotion and range of motion training of the shoulder.

A shortage of physiotherapist (PT) manpower is a barrier for providing better rehabilitation service in Hong Kong. Quality training can benefit patients with better recovery, on the contrary, insufficient training may cause a longer length of stay, readmission, and thus the burden of healthcare system. The estimated cost for PT services in Hospital Authority was HK$7.0 Billion in 2020. A novel Danish robot with a 7-joint robotic arm became popular in Denmark and Germany in the last two years. The robot is designed for lower limb patient rehabilitation. It can enhance the mobility of patients. Based on the experience of a university hospital in Denmark, this robotic rehabilitation was well accepted by both patients and PTs. Function-wise, the robot provides many clinical benefits to patients, especially stroke ones. A physiotherapist’s time can be saved when the robot is being used. The cost-effectiveness of ROBERT${}^{\text{®}}$ is better than PT performing repetitive exercises for lower limbs. The robot potentially provides a cost-effective solution to the Hong Kong healthcare system.

Magnetic wheels are a powerful solution to design inspection climbing robots with excellent mobility. Magnetic wheels optimization based on simulations and the results that were obtained on prototypes are presented. The measured adhesion was doubled between the classic configuration and a novel multilayer one sharing exactly the same four magnets and the same total volume of iron. This know-how is then applied to optimize magnetic wheels for the existing robot called MagneBike. The adhesion force has been multiplied by 2 to 3 times depending on the conditions. Those amazing improvements open new possibilities for miniaturization of climbing robots or payload's increase.

In this paper, the artificial intelligence control algorithm for steering robot of steering wheel is studied. The steering movement of wheeled soccer robot is controlled by artificial intelligence control algorithm, and the steering movement is modeled and simulated. Firstly, the characteristics of artificial neurons are simulated and a similar control model is constructed to complete the simulation of football. The artificial intelligence control algorithm has a dynamic feedback item compared with the traditional intelligent model, which has a better effect on the steering control of the wheeled soccer robot. In this paper, artificial intelligence control algorithm is used to optimize the parameters of artificial intelligence control algorithm, and the output of control signal of each steering part of wheeled soccer robot is simulated in the experiment, and the control of the steering action of wheeled soccer robot by artificial intelligence control algorithm is verified by experiments. Then the artificial intelligence control algorithm forms the connection structure. This method provides a good reference for steering control of wheeled soccer robots.

This paper focuses on the design of badminton robots, and designs high-precision binocular stereo vision synchronous acquisition system hardware and multithreaded acquisition programs to ensure the left and right camera exposure synchronization and timely reading of data. Aiming at specific weak moving targets, a shape-based Brown motion model based on dynamic threshold adjustment based on singular value decomposition is proposed, and a discriminative threshold is set according to the similarity between the background and the foreground to improve detection accuracy. The three-dimensional trajectory points are extended by Kalman filter and the kinematics equation of badminton is established. The parameters of the kinematics equation of badminton are solved by the method of least squares. Based on the fractal Brownian motion algorithm, a real-time robot pose estimation algorithm is proposed to realize the real-time accurate pose estimation of the robot. A PID control model for the badminton robot executive mechanism is established between the omnidirectional wheel speed and the robot’s translation and rotation movements to achieve the precise movement of the badminton robot. All the algorithms can meet the system’s requirements for real-time performance, realize the badminton robot’s simple hit to the ball, and prospect the future research direction.

Spastic hypertonia causes loss of range of motion (ROM) and contractures in patients with post-stroke hemiparesis. The pronation/supination of the forearm is an essential functional movement in daily activities. We developed a special module for a shoulder-elbow rehabilitation robot for the reduction and biomechanical assessment of pronator/supinator hypertonia of the forearm. The module consisted of a rotational drum driven by an AC servo motor and equipped with an encoder and a custom-made torque sensor. By properly switching the control algorithm between position control and torque control, a hybrid controller able to mimic a therapist’s manual stretching movements was designed. Nine stroke patients were recruited to validate the functions of the module. The results showed that the affected forearms had significant increases in the ROM after five cycles of stretching. Both the passive ROM and the average stiffness were highly correlated to the spasticity of the forearm flexor muscles as measured using the Modified Ashworth Scale (MAS). With the custom-made module and controller, this upper-limb rehabilitation robot may be able to aid physical therapists to reduce hypertonia and quantify biomechanical properties of the muscles for forearm rotation in stroke patients.

A set of mobile robots is placed at arbitrary points of an infinite line. The robots are equipped with GPS devices and they may communicate their positions on the line to a central authority. The collection contains an unknown subset of “spies”, i.e., byzantine robots, which are indistinguishable from the non-faulty ones. The set of the non-faulty robots needs to rendezvous in the shortest possible time in order to perform some task, while the byzantine robots may try to delay their rendezvous for as long as possible. The problem facing a central authority is to determine trajectories for all robots so as to minimize the time until all the non-faulty robots have met. The trajectories must be determined without knowledge of which robots are faulty. Our goal is to minimize the competitive ratio between the time required to achieve the first rendezvous of the non-faulty robots and the time required for such a rendezvous to occur under the assumption that the faulty robots are known at the start.

In this paper, we give rendezvous algorithms with bounded competitive ratio, where the central authority is informed only of the set of initial robot positions, without knowing which ones or how many of them are faulty. In general, regardless of the number of faults $f\le n-2$ it can be shown that there is an algorithm with bounded competitive ratio. Further, we are able to give a rendezvous algorithm with optimal competitive ratio provided that the number $f$ of faults is strictly less than $\lceil n/2\rceil$. Note, however, that in general this algorithm does not give an estimate on the actual value of the competitive ratio. However, when an upper bound on the number of byzantine robots is known to the central authority, we can provide algorithms with constant competitive ratios and in some instances we are able to show that these algorithms are optimal. Moreover, in the cases where the number of faults is either $f=1$ or $f=2$ we are able to compute the competitive ratio of an optimal rendezvous algorithm, for a small number of robots.

Human-robot interaction studies up to now have been limited to simple tasks such as route guidance or playing simple games. With the advance in robotic technologies, we are now at the stage to explore requirements for highly complicated tasks such as having human-like conversations. When robots start to play advanced roles in our lives such as in health care, attributes such as trust, reliance and persuasiveness will also be important. In this paper, we examine how the appearance of robots affects people's attitudes toward them. Past studies have shown that the appearance of robots is one of the elements that influences people's behavior. However, it is still unknown what effect appearance has when having serious conversations that require high-level activity. Participants were asked to have a discussion with tele-operated robots of various appearances such as an android with high similarity to a human or a humanoid robot that has human-like body parts. Through the discussion, the tele-operator tried to persuade the participants. We examined how appearance affects robots' persuasiveness as well as people's behavior and impression of robots. A possible contribution to machine consciousness research is also discussed.

A stereovision algorithm is proposed for visual odometry to estimate motion of mobile robot by providing feature pair sequence. It is composed of feature extracting, matching and tracking. Firstly, corners are extracted as features by Harris operator and grid-based optimizing. In feature matching and tracking, serious problems are caused by variable illumination between stereo images. An improved Moravec's Normalized Cross Correlation (MNCC) algorithm is presented to reduce illumination affect in computing correspondence of corners. On current stereo image pair, extracted corners are matched by correlation-based bidirectional algorithm and outliers are rejected by epipolar constraint. Matched corners are tracked in pre-estimated search windows. The computational cost is greatly reduced by limiting number of corners, pre-estimating search window and feature local-updating. Simulation results validate that our algorithm is efficient and reliable.